|
|
|
|
|||||
|
||||||
From: Herbert Prinz (no email)
Date: Tue Nov 02 2004 - 02:24:49 EST
Dear Alexandre,
In a message from 2004-10-09 I had claimed that it is mathematically
wrong to average the observed altitudes before feeding them into the
sight reduction process, because the only rigorously correct method
minimizes the sum of the squares of the altitude residuals (i.e.
observed altitude minus computed altitude).
You had asked for the references that I made to articles discussing
relevant LSQ algorithms in
"Navigation: Journal of The Institute of Navigation"
C. DeWit, Optimal Estimation of a Multi-Star Fix, Vol. 21, No 4, 1974-76
Robert W. Severance, Overdetermined Celestial Fix by Iteration, Vol. 36,
No 4, Winter 1989-90
Thomas R. Metcalf, Frederic T. Metcalf, On the Overdetermined Celestial
Fix, Vol. 38, No. 1, Spring 1991
There are probably many more, but these are the references in have noted
down.
Regards
Herbert Prinz
|